Cannot load 2 bit quantized ggml model on Windows

[dillfrescott]

C:\Users\micro\Downloads>main -m ggml-model-q2_0.bin
main: seed = 1681700481
llama.cpp: loading model from ggml-model-q2_0.bin
error loading model: unrecognized tensor type 5

llama_init_from_file: failed to load model
main: error: failed to load model 'ggml-model-q2_0.bin'

[rabidcopy]

#1004 Not merged yet and still a WIP. You will need to use this PR to load models quantized to 2-bit.